Zengjie Fanab,
Bin Liuc,
Zhangpeng Lia,
Limin Maa,
Jinqing Wang*a and
Shengrong Yanga
aState Key Laboratory of Solid Lubrication, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, Lanzhou 730000, P. R. China. E-mail: jqwang@licp.cas.cn; Fax: +86 931 4968076; Tel: +86 931 4968076
bUniversity of Chinese Academy of Sciences, Beijing 100080, P. R. China
cSchool of Stomatology, Lanzhou University, Lanzhou 730000, P. R. China
First published on 8th April 2014
CuO microstructures with urchin-like, flower-like and sheet-like morphologies were directly grown on a Ni foam via a simple and low-cost hydrothermal method. The aim was to construct a three-dimensional porous hybrid electrode for an amperometric non-enzymatic glucose sensor. As a result, the as-prepared hybrid electrode with a flower-like morphology exhibited a higher electrocatalytic activity towards the oxidation of glucose compared to electrodes with other morphologies and a pristine Ni foam electrode. The flower-like CuO/Ni foam electrode displayed a high sensitivity of 1084 μA mM−1 cm−2 to glucose ranging from 0.5 μM to 3.5 mM, which is higher than most of the reported CuO based electrodes, and also a low detection limit of 0.16 μM (signal/noise = 3). Notably, poisoning by chloride ions and interference from ascorbic acid, uric acid, dopamine, and sucrose were negligible. These results indicate that the flower-like CuO/Ni foam hybrid electrode is a promising candidate for amperometric non-enzymatic glucose detection.
Noble metals and their alloys, such as Pt, Au, Pd, Pt–Pb, Pt–Ru and Pt–Au, are widely reported as electrode materials for constructing non-enzymatic glucose sensors.9–11 However, the sensors based on these materials have drawbacks such as high cost of rare metal precursors, a narrow linear range, poor selectivity and low immunity to chloride ions.3,11–13 Surprisingly, CuO with a flower-like morphology shows high sensitivity, good electrochemical activity, and the possibility of promoting electron transfer reactions at a lower overpotential.13–15 Therefore, CuO has been considered as an ideal substitute for noble metals and their alloys.16 However, its complicated synthesis, high cost and multi-step surface immobilization process limit its application as an electrode material.15,17 In addition, this material cannot be in contact with the electrolyte because aggregation occurs on the electrode surface which is caused by the high surface energy, leading to a significant deterioration in the catalytic performance.18 Therefore, it is necessary to seek a facile and low-cost procedure for synthesizing flower-like CuO microstructures with high catalytic performance for practical application.
Recently, sensors with porous structures have attracted great research interest because of rapid electrochemical reactions and fast mass transport of ions.19 As a three-dimensional (3D) porous structure, Ni foam has been widely applied as a supporting scaffold for constructing hybrid electrodes.20,21 The porous features enable good access of ions and electrons to the active surfaces, herein enhancing the electrochemical performance.22 Moreover, many studies have found that hybrid electrodes, such as Cu/Ni,23,24 CuO/TiO2,25 and CuO/Co3O4 (ref. 26) possess better performance because of synergetic effects compared with single metal or oxide electrodes. To utilize the advantages of flower-like CuO and Ni foam, in this paper, a hybrid electrode, CuO microflower/Ni foam, has been built and applied as a non-enzymatic glucose sensor. The as-prepared hybrid electrode is low-cost, easy to mass produce and simple to fabricate for practical applications. More importantly, it shows a higher sensitivity than most reported electrodes based on CuO, and resistance against poisoning by chloride ions.
![]() | ||
Fig. 1 XRD patterns of CuO microsheet/Ni foam, CuO microflower/Ni foam, CuO microurchin/Ni foam, and the standard data for CuO as a reference (JPCDS no. 78-0428). |
The possible chemical reaction for the formation of CuO can be depicted as follows:27
CO(NH2)2 + H2O → 2NH3 + CO2 | (1) |
NH3 + H2O ![]() | (2) |
Cu2+ + 2OH− → Cu(OH)2 | (3) |
![]() | (4) |
In the suggested reaction scheme, urea played an important role in the formation of various CuO microstructures. As a good precipitating agent, urea can hydrolyze in the aqueous solution to produce ammonia and carbon dioxide [eqn (1)]. The formed ammonia further reacts with the aqueous solution and produces ammonium ions and OH− ions [eqn (2)]. Therefore, eqn (2)is a crucial chemical reaction because this reaction can effectively control the supply of OH− ions which are most important for the formation of different CuO microstructures.28 Since OH− ions are derived from urea, we can control the supply of OH− ions by simply changing the amount of urea used in the reaction to produce CuO with different morphologies. Moreover, the OH− ions produced from eqn (2) can further react with copper ions and form the copper hydroxide nuclei [eqn (3)]. These can be used as building blocks for constructing different morphologies.29 At 90 °C, the formed copper hydroxide will decompose into copper oxide [eqn (4)].
The morphology of the pristine Ni foam and the various microstructures of CuO/Ni foam were characterized by SEM. As clearly shown in Fig. S1A,† the smooth Ni foam has a 3D porous and cross-linked grid structure. After various microstructures of CuO grew on the Ni foam, the surface of the Ni foam became more coarse. Meanwhile, the porous structure of the Ni foam remained even after reaction in strong alkaline ammonia hydroxide solution (Fig. S1B–D†). This special Ni foam structure would not only reduce the diffusion resistance of the electrolyte but also enhance the ion transmission performance.22 Fig. 2a and b display the SEM images of CuO microurchin/Ni foam synthesized with addition of a small quantity of urea. It can be found that urchin-like microspheres, with a diameter of ∼1.5 μm, densely cover the backbone of the Ni foam. The enlarged image in Fig. 2b indicates that the urchin-like microspheres are composed of many tiny nanoparticles, which radiate from the center of CuO and form an urchin-shaped spherical structure. These urchin-like microspheres were further examined by TEM (Fig. 2c) and high-resolution TEM (HRTEM) (Fig. 2d and f). The TEM image demonstrates that the urchin-like microspheres consist of lots of tiny nanoparticles. The HRTEM image of a single nanoparticle reveals that the lattice fringe spacing is 0.21 nm, which is close to the (200) CuO plane. Fig. 2e provides a typical select area electron diffraction (SAED) pattern taken of the nanoparticles, in which the primary ring pattern is induced by the polycrystal. Increasing the amounts of urea, the morphology of CuO changed from a microurchin to a microflower structure. The SEM images of CuO microflower/Ni foam with different magnifications are shown in Fig. 3a and b. Many CuO particles, with a diameter of ∼1.8 μm, densely cover the surface of the Ni foam. From a close observation, it is interesting to see that each particle possesses a 3D flower-like morphology (Fig. 3b). The TEM image of an individual “petal” indicates that it has a nanosheet-like structure (Fig. 3c). Fig. 3d shows a typical HRTEM image, and the lattice fringe spacing is determined to be 0.15 nm, corresponding to the (220) CuO lattice. In addition, the feature of the CuO polycrystal can also be demonstrated by SAED analysis. Upon further increasing the amount of urea, the morphology of CuO obtained is quite different from the former morphologies. As can be seen from Fig. 4a and b, upright rectangular sheet-like CuO grows on the backbone of the Ni foam, and some tiny spherical particles are sparsely located between the sheets. These particles may be the incompletely dissolved flower-like CuO. The reason for this is that the formed Cu(OH)2 is gradually dissolved when it reacts with excess ammonium hydroxide to form a [Cu(NH3)4]2+ solution. The reaction can be illustrated as follows:27
Cu(OH)2 + 4NH3 + 2NH4+ → [Cu(NH3)4]2+ + 2NH3·H2O |
![]() | ||
Fig. 2 SEM images (a) and (b), TEM and HRTEM images (c), (d) and (f), and SAED image (e) of CuO microurchin/Ni foam. |
![]() | ||
Fig. 3 SEM images (a) and (b), TEM and HRTEM images (c), (d) and (f), and SAED image (e) of CuO microflower/Ni foam. |
![]() | ||
Fig. 4 SEM images (a) and (b), TEM and HRTEM images (c), (d) and (f), and SAED image (e) of CuO microsheet/Ni foam. |
The sheet-like CuO can be further characterized by TEM (Fig. 4c). The lattice fringe spacing is determined to be 0.21 nm, which is consistent with the (200) CuO plane (Fig. 4d and f). Moreover, the relevant SAED pattern of CuO also belongs to the polycrystal (Fig. 4e).
In terms of a glucose sensor, a good amperometric response to glucose concentration is a crucial evaluation criterion to determine whether it can be used for glucose analysis. Therefore, the amperometric response of the three types of electrode to different glucose concentrations was tested and used as the evaluation criteria for choosing the optimal sensor. As shown in Fig. 5b, the CuO microflower/Ni foam electrode yields a much larger current response than the other electrodes upon successive addition of glucose. Although the CuO microsheet/Ni foam electrode has a larger redox peak current than the CuO microurchin/Ni foam electrode (Fig. 5a), its amperometric response is much smaller, indicating a poorer electrocatalytic performance for glucose. Therefore, we chose the CuO microflower/Ni foam electrode as the experimental group and pristine Ni foam electrode as the control group for studying the degree of improvement of the electrochemical performance after surface modification with flower-like CuO.
The electrochemical properties of the CuO microflower/Ni foam electrode and Ni foam electrode were investigated and compared using CV measurements in a 0.1 M NaOH solution, in the absence and presence of glucose. As can be seen from Fig. 6a, no obvious redox peak appears in the curve A of Ni foam electrode. While for the hybrid electrode, a pair of broad and negative shifts of the redox peak with higher current intensity can be observed in the curve E. The negatively shifted potential may be attributed to the formation of flower-like CuO microstructures on the Ni foam, which can speed up electron transfer.31 Upon the addition of various concentrations of glucose in the 0.1 M NaOH solution, only a slight current response can be detected for the Ni foam (curves B–E in Fig. 6a and Fig. S3B–E†), while a strong and gradual current response increase to glucose occurs on the as-prepared CuO microflower/Ni foam hybrid electrode (curves F–H in Fig. 6a). In particular, the oxidation peak current in the presence of 0.2 mM glucose is approximately two times larger than in the absence of glucose, indicating that the incorporation of CuO microflowers onto the Ni foam can greatly enhance the electrocatalytic activity of the electrode. As the catalytic center, high-density nanoscale “petals” not only provide large surface areas and structural defects, but also avoid the internal resistance aroused by the high aspect ratio, which can enhance the electrocatalytic performance of the electrode.15 The exact mechanism for the oxidation of glucose in alkaline media on the CuO electrode is still unclear; however, a possible mechanism involving two stages was put forward by previous works, and can be described as follows:32,33
Firstly, CuO can be electrochemically oxidized into Cu(III) species, such as CuOOH or Cu(OH)4−, which are strong oxidizing agents.
CuO + OH− → CuOOH or CuO + H2O + 2OH− → Cu(OH)4− + e− |
Secondly, glucose is enolized, after de-protonation and isomerization in an alkaline medium, which can be oxidized to gluconolactone when it reacts with Cu(III), and then is further hydrolyzed to gluconic acid.
Cu(III) + glucose → gluconolactone + Cu(II) |
Gluconolactone → gluconic acid (hydrolysis) |
Fig. 6b shows the amperometric response of the CuO microflower/Ni foam electrode to the successive addition of 100 mM glucose under a potential window ranging from +0.3 to +0.5 V. The current response increases upon improving the applied potential. When the applied potential reaches +0.5 V, the current response becomes almost steady (Fig. S4†). Moreover, a high potential can oxidize some interfering species, such as AA, UA, and DA, causing unnecessary interfering signals. Thus, +0.5 V was chosen as the optimal potential for amperometric glucose sensing.
The amperometric responses of the Ni foam electrode and CuO microflower/Ni foam electrode to the successive addition of 2 mM glucose at a constant potential of +0.5 V in a 0.1 M NaOH solution are compared in Fig. S5.† Obviously, a step-like current change can be observed from the hybrid electrode, while no current response appears for the Ni foam electrode, demonstrating that the performance of the CuO microflower/Ni foam hybrid electrode outperforms the pristine Ni foam electrode. At a constant potential of +0.5 V, the amperometric responses of the CuO microflower/Ni foam electrode to the successive addition of 2 mM glucose are depicted in Fig. 7a. Apparently, the current response of the hybrid electrode exhibits a linear dependence on glucose concentration. The calibration curve for the sensor is shown in Fig. 7b, revealing that the sensor displays a linear response ranging from 0.5 μM to 3.5 mM (correlation coefficient was 0.992) with a sensitivity of 1084 μA mM−1 cm−2 (the calculation of electrode area is the first step to determine the sensitivity of this sensor. The detailed calculation procedures of the electrode area are given in the ESI.†) and a detection limit of 0.16 μM (signal/noise = 3). By contrast, the sensitivity of our glucose sensor is higher than most previously reported (see Table S3†) and the reason for this can be explained as follows.34–38 In the as-prepared hybrid electrode, the CuO microflower has a larger surface area and can be used as the active center, which can promote electron transfer and enhance the electrocatalytic performance.39 Meanwhile, the porous Ni foam enabled each CuO microflower to access the electrolyte more easily, thereby contributing to bringing the electrocatalytic performance of each nanoparticle into full play. Namely, the synergetic effect of the CuO microflower and the Ni foam played the important role in jointly enhancing the electrocatalytic performances of this hybrid electrode.
![]() | ||
Fig. 7 (a) The response of the CuO microflower/Ni foam electrode to the successive addition of glucose from 5 μm to 3.5 mM. (b) The calibration curve of the CuO microflower/Ni foam electrode. |
It is crucial for the sensor to have excellent anti-interference properties, because some interferents, such as UA, AA, DA, sucrose, and other compounds in the real serum sample, can cause interfering signals and influence the test result.40 As shown in Fig. 8a, all of the potential interferents cause almost a negligible amperometric response on the CuO microflower/Ni foam electrode. On the other hand, the non-enzymatic glucose sensors based on metals or their oxides are usually poisoned by chloride ions.30 Therefore, the effect of chloride ions on the electroanalysis of glucose on this electrode was also examined with the addition of 200 mM NaCl in 0.1 M NaOH solution. There is no obvious current change observed, demonstrating that the electrode also can be used in the presence of a high concentration of chloride ions.
![]() | ||
Fig. 8 (a) Effect of interferents (0.01 mM DA, 0.01 mM UA, 0.01 mM AA, 0.01 mM sucrose, and 200 mM NaCl). (b) Stability of the sensor stored at ambient conditions over two weeks. |
The stability of the developed sensor was further examined by measuring its current response to glucose over two weeks, during which only a 10% loss in the current signal was observed, showing the sensor has good stability (Fig. 8b). The current responses for 1 mM glucose were measured 8 times using the same electrode and only a relative standard deviation (RSD) of 5.4% is presented. The reproducibility was investigated, six identical electrodes were made, with a RSD of 4.2%, illustrating the reliability of this method.
In order to verify the performance of the prepared sensor in practical applications, it was used to detect glucose in human blood serum samples of diabetic and healthy people, which were obtained from a hospital. The detailed testing process can be found in our previous report.32 The recovery of glucose was determined by the standard addition of pure glucose to the serum samples and the corresponding results are shown in Table S4.† It can be observed that the sensor gives recoveries in the range of 92–106%, suggesting its good quantitative accuracy.
Footnote |
† Electronic supplementary information (ESI) available: Reactive reagents (Table S1), the surface area of CuO with different morphologies (Table S2), SEM images of Ni foam, CuO microurchin/Ni foam, CuO microflower/Ni foam and CuO microsheet/Ni foam (Fig. S1), CV curves of Ni foam electrode, CuO microflower/Ni foam, flat Ni plate and CuO microflower/Ni plate (Fig. S2), CV curves of Ni foam electrode (Fig. S3), the effect of potential on amperometric response (Fig. S4), the amperometric response of Ni foam electrode (Fig. S5), the performance comparison between CuNWs/GTE and the other reported glucose sensors (Table S3), and the amperometric determination of glucose in human blood serum samples (Table S4). See DOI: 10.1039/c3ra47422d |
This journal is © The Royal Society of Chemistry 2014 |